TPS25983: Inquiry about the TPS259830ONRGE (Circuit breaker)

Hi Arush,
unfortunately we were wrong, in the tests we used a power supply with a very slow response from empty to load and this deceived us, that's why the e-fuse did not enter protection with only 1nF. Using a power supply with reactive response the e-fuse returned to protection in the presence of load.
Reviewing the data sheet and the electronic sheet, unfortunately with a Vin of 15 V, a total output capacity of 1.4mF (in the diagram they are not visible because they are at the output of the boost converter connected to the output of the e-fuse, which even if turned off charges the capacitors via the diode) a load resistance of 5.3 ohm (maximum load condition) and a CdVdt og 10nF it is clear that we are outside of Thermal Shutdown Plot During Start-up (Figure 8-2. data sheet). Is it possible to do something?

Best Regards,

Maurizio

Hi Maurizio,

Please confirm if my understanding is correct.

The output cap of 1.4mF will be charged through output boost converter (through the diodes) so it won't cause inrush current through the eFuse. Any scenario in which it will cause inrush current.

If 1.4mF is not gonna draw current, then reducing dvdt cap to 2nF or under should allow startup with 5Ohms load. In case of resistive load, faster startup is better. In case of capacitive load, slower startup to reduce inrush current is better. 

https://www.ti.com/tool/download/TPS25983-CALC This is design calculator. This can help in deciding dvdt cap value based on voltage and load (resistive and capacitive).

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1368578/faq-tps25983-how-to-set-dvdt-cap-and-auto-retries-for-very-large-output-capacitor This is FAQ on setting dvdt for large output caps.

Best Regards,
Arush

Hi Harush,

For greater clarity, I send you a simplified diagram of the converter where we used the e-fuse. The problem is essentially due to the presence of the load at startup. At power on as mentioned the boost is off, we are using a CdVdt of 4.7nF, it would seem the best compromise at the input voltage of 15V, but the minimum RL must not fall below 10 ohm. We would like to know if it is necessary to use a TI e-fuse with a higher Thermal Shutdown During Start-up in these conditions? If so, what type do you suggest? I also thought of disconnecting RL during the start up phase with a hi side mosfet at the output... Thank you very much for the precious advice you are giving me.

Best Regards,

Maurizio

Hi Maurizio,

Thank you for providing the block diagram. Using a FET to disconnect the load during startup and then based on PG signal connecting the load should work in your application.

The eFuse performs the thermal shutdown when the load is high during startup so that SOA of the device is maintained. 

Best Regards,
Arush

Hi Arush,

yes i also think this is the solution with the least risk of load related problems.

Thanks a lot for the quick answers.

Best Regards,

Maurizio